Over the past 20 years, numerical skills have become increasingly necessary for navigating the modern health care environment. Health professionals often provide numerical information about benefits and risks of medical interventions and lifestyles, and direct-to-consumer advertising often summarizes changes in incidence of various diseases and describes the effects of screenings and novel drugs. Unfortunately, many people struggle to grasp numerical concepts that are essential for understanding health-relevant information. Even highly educated individuals tend to have difficulties interpreting and using a host of elementary probability expressions. In short, the general public lacks basic numeracy, which limits their risk literacy - the ability to accurately interpret and make good decisions based on information about risk (see www.RiskLiteracy.org). In this talk, I will present a collection of studies investigating three related research questions. This research involves more than 5000 participants from 30 countries (e.g., China, England, Japan, India, Pakistan, Spain, Sweden, and the United States) and diverse walks of life (e.g., medical professionals, patients, general populations, web panels).
First, I will present research investigating how helpful numbers are when communicating health-relevant risks. Our research converges to suggest that a significant proportion of the population has problems understanding even very simple numerical expressions of probability about health. This research also shows that numeracy predicts performance across paradigmatic judgment and medical decision making tasks.
Second, I will present research investigating whether numeracy predicts health outcomes. Our research shows that compared to patients with high numeracy, patients with low numeracy showed higher prevalence of comorbidity (e.g., myocardial infarction, COPD, or HIV/AIDS) and tended to take more prescribed medications. Of note, these conclusions hold after controlling for the effect of patients’ demographics, risky habits, BMI, levels of health literacy, trust in physicians, and satisfaction with role in medical decision making, suggesting that numeracy is uniquely related to important health outcomes.
Transparent visual aids are simple graphical representations of numerical expressions of probability about health and can confer benefits when communicating risk information. In the last part of this talk, I will present the results of several interventions designed to improve risk communication in people with low numeracy. These intervention studies indicated that well-designed visual aids (1) help individuals with low in numeracy make more accurate assessments of information about treatment risk reduction, (2) reduce biased medical decision making, (3) increase accuracy of diagnostic inferences beyond the effect of other transparent information formats (natural frequencies), and (4) help promote shared decision making and healthy behaviors (e.g., condom use and screening for STDs).
I will conclude that appropriately designed visual aids can be highly effective, transparent, and ethically desirable tools for improving risk communication, limiting the influence of different levels of numeracy. Theoretical mechanisms, open questions, and emerging applications of our research will be discussed.